Individual differences in phoneme categorization Effie Kapnoula, - - PowerPoint PPT Presentation

Individual differences in phoneme categorization

Effie Kapnoula, Bob McMurray, Eunjong Kong, Matthew Winn, & Jan Edwards

19th Mid-Continental Phonetics & Phonology Conference

The problem of lack of invariance

• There is no one-to-one relation between a sound (i.e. formant

frequencies) and the perceived phoneme

Hillenbrand, Getty, Clark & Wheeler, 1995

2

The problem of lack of invariance

• There is no one-to-one relation between a sound (i.e. formant

frequencies) and the perceived phoneme

• One solution: categorical perception

3

The problem of lack of invariance

• There is no one-to-one relation between a sound (i.e. formant

frequencies) and the perceived phoneme

• One solution: categorical perception

bar

par

1 2 3 4 5 6 7 par bar

4

The problem of lack of invariance

• There is no one-to-one relation between a sound (i.e. formant

frequencies) and the perceived phoneme

• One solution: categorical perception

+Simple solution +Fast commitment

bar

par

1 2 3 4 5 6 7 par bar

5

Two alternative forced choice (2AFC)

Werker & Tees, 1987; Joanisse et al, 2000; López-Zamora et al, 2010

6

The problem of lack of invariance

• There is no one-to-one relation between a sound (i.e. formant

transitions) and the perceived phoneme

• One solution: categorical perception

+Simple solution +Fast commitment

• Alternative: gradient perception

bar

par

1 2 3 4 5 6 7 par bar

bar

par

1 2 3 4 5 6 7 par bar

7

The problem of lack of invariance

• There is no one-to-one relation between a sound (i.e. formant

transitions) and the perceived phoneme

• One solution: categorical perception

+Simple solution +Fast commitment

• Alternative: gradient perception

+Flexibility +Late commitment +Keep useful within-category information

bar

par

1 2 3 4 5 6 7 par bar

bar

par

1 2 3 4 5 6 7 par bar

8

Gradiency in speech perception

5 10 15 20 25 30 35 40 0.02 0.03 0.04 0.05 0.06 0.07 0.08

VOT (ms)

Category Boundary

Response = Response = Looks to Looks to Competitor Fixations

McMurray, Tanenhaus & Aslin (2002)

• Evidence for gradiency from eye-movements

9

Two alternative forced choice (2AFC)

• Is gradiency good or bad for speech perception?

Werker & Tees, 1987; Joanisse et al, 2000; López-Zamora et al, 2010

?

10

Gradiency in speech perception

bull pull bull

• Measuring gradiency: Visual analog scaling (VAS) task

11

Gradiency in speech perception

Kong, E. J., & Edwards, J. (2011)

12

Gradiency in speech perception

Kong, E. J., & Edwards, J. (2011)

13

• Summary points:

 Listeners are capable of gradient categorization of phonemes  The VAS task allows for this gradiency to be expressed in participants’ responses

Summary and aims

14

• Summary points:

 Listeners are capable of gradient categorization of phonemes  The VAS task allows for this gradiency to be expressed in participants’ responses

• Where does gradiency come from? Is it good or bad for speech perception?

Summary and aims

15

• Summary points:

 Listeners are capable of gradient categorization of phonemes  The VAS task allows for this gradiency to be expressed in participants’ responses

• Where does gradiency come from? Is it good or bad for speech perception?

 Establish a way of quantifying gradiency via the VAS task

Summary and aims

16

• Summary points:

 Listeners are capable of gradient categorization of phonemes  The VAS task allows for this gradiency to be expressed in participants’ responses

• Where does gradiency come from? Is it good or bad for speech perception?

 Establish a way of quantifying gradiency via the VAS task 1. Investigate possible sources of gradiency (e.g. executive function) 2. Link gradiency to multiple cue use 3. Examine whether gradiency is good or bad for speech perception

Summary and aims

17

• Stimuli:
• Seven (7) VOT steps (primary cue) and five (5) F0 steps (secondary cue)

Method

labial alveolar Real words bull-pull den-ten Nonwords buv-puv dev-tev CVs buh-puh deh-teh

F0 steps 5 4 3 2 1 1 2 3 4 5 6 7 VOT steps

18

• Stimuli:
• Seven (7) VOT steps (primary cue) and five (5) F0 steps (secondary cue)
• Tasks:
• Visual analog scaling (VAS) task
• Two alternative forced choice (2AFC)

Method

labial alveolar Real words bull-pull den-ten Nonwords buv-puv dev-tev CVs buh-puh deh-teh bull pull pull bull

19

• Additional tasks:

 Trail making task (cognitive flexibility)  N-Back task (working memory)  Flanker task (inhibition)

Method

non-speech cognitive processes

20

• Additional tasks:

 Trail making task (cognitive flexibility)  N-Back task (working memory)  Flanker task (inhibition)  AZ-bio (sentences in babbling noise - 1:1 STN ratio)

Method

non-speech cognitive processes

21

• Additional tasks:

 Trail making task (cognitive flexibility)  N-Back task (working memory)  Flanker task (inhibition)  AZ-bio (sentences in babbling noise - 1:1 STN ratio)

• Participants: 130 undergraduates at the U of Iowa

Method

non-speech cognitive processes

22

Results

23

Results

10 20 30 40 0 10 20 30 40 50 60 70 80 90100 10 20 30 40 0 10 20 30 40 50 60 70 80 90100 10 20 30 40 0 10 20 30 40 50 60 70 80 90100 10 20 30 40 0 10 20 30 40 50 60 70 80 90100

Sub 8 Sub 9 Sub 7 Sub 68

24

Results: Quantifying gradiency

25

Results: Quantifying gradiency

• Extracting gradiency from VAS data

F0 steps 5 4 3 2 1 1 2 3 4 5 6 7 VOT steps F0 steps 5 4 3 2 1 1 2 3 4 5 6 7 VOT steps VOT steps F0 steps θ s VOT steps F0 steps θ s

26

Results: Quantifying gradiency

• Extracting gradiency from VAS data

F0 steps 5 4 3 2 1 1 2 3 4 5 6 7 VOT steps F0 steps 5 4 3 2 1 1 2 3 4 5 6 7 VOT steps VOT steps F0 steps θ s VOT steps F0 steps θ s

Steep s slope Shallow s slope

gradient categorical

27

Results: Quantifying secondary cue use

• Extracting F0 use from 2AFC data

0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 b 2AFC response p VOT step 1 2 F0 = 90 hZ F0 = 125 hZ

28

Results

29

Results: Stimulus and place effects

20 40 60 80 100 1 2 3 4 5 6 7 b VAS response p VOT step NP NW RW 0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 b 2AFC response p VOT step Real words 0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 b 2AFC response p VOT step

Nonwords

0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 b 2AFC response p VOT step

CVs 90hZ 125hZ

20 40 60 80 100 1 2 3 4 5 6 7 b VAS response p VOT step alv lab

30

Results: Stimulus and place effects

0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 b 2AFC response p VOT step Real words 0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 b 2AFC response p VOT step

Nonwords

0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 b 2AFC response p VOT step

CVs 90hZ 125hZ

20 40 60 80 100 1 2 3 4 5 6 7 b VAS response p VOT step NP NW RW 20 40 60 80 100 1 2 3 4 5 6 7 b VAS response p VOT step alv lab

F<1 F<1

31

Results: Place differences in F0 use

F(1,250) = 27.8, p < 0.001

0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 b 2AFC response p VOT step

Labials

0.2 0.4 0.6 0.8 1 1 2 3 4 5 6 7 VOT step

Alveolars

1 5

90hZ 125hZ

32

Results

1.

Do individual differences in gradiency derive from differences in general cognitive function?

33

Results

1.

Do individual differences in gradiency derive from differences in general cognitive function?

 EF measures did not account for a statistically significant amount of variance in VAS slope, F(3,108)=1.75, p=.162, or F0 use, F<0

gradient categorical gradient categorical gradient categorical

34

Results

1.

Do individual differences in gradiency derive from differences in general cognitive function?

 EF measures did not account for a statistically significant amount of variance in VAS slope, F(3,108)=1.75, p=.162, or F0 use, F<0  Speech perception processes may be played out on a different level of processing than higher cognitive processes, such as working memory

35

Results

1.

Do individual differences in gradiency derive from differences in general cognitive function?

2.

Are individual differences in gradiency linked to multiple cue use?

36

Results

1.

Do individual differences in gradiency derive from differences in general cognitive function?

2.

Are individual differences in gradiency linked to multiple cue use?

 Positive relationship: Better encoding of fine-grained detail (more gradiency) enables access to multiple cues  Negative relationship: Listeners who use more cues have more accurate, sharper boundaries

37

Results

β=-0.305, t=-3.4, p < 0.01

gradient categorical

38

Results

1.

Do individual differences in gradiency derive from differences in general cognitive function?

2.

Are individual differences in gradiency linked to multiple cue use?

 Positive relationship: Better encoding of fine-grained detail (more gradiency) enables access to multiple cues  Negative relationship: Listeners who use more cues have more accurate, sharper boundaries

39

Results

1.

Do individual differences in gradiency derive from differences in general cognitive function?

2.

Are individual differences in gradiency linked to multiple cue use?

3.

In what way are these differences important for speech perception?

40

Results

• Gradiency and perception of speech-in-noise

r = .164, p=.068

gradient categorical

41

Results

• Gradiency and perception of speech-in-noise

r = .243, p=.007 r = .164, p=.068

gradient categorical

42

Results

• Gradiency and perception of speech-in-noise

Gradiency Speech-in- noise Gradiency Speech-in- noise Working Memory 1) 2)

43

Results

• Gradiency and perception of speech-in-noise

R2 = 0.019

β=-0.14, t=-1.48, p = .143

categorical gradient

44

Results

• Gradiency and perception of speech-in-noise

Gradiency Speech-in- noise Gradiency Speech-in- noise Working Memory 1) 2)

45

Results

1.

Do individual differences in gradiency derive from differences in general cognitive function?

2.

Are individual differences in gradiency linked to multiple cue use?

3.

In what way are these differences important for speech perception?

 More gradient listeners tend to better perceive speech in noise

46

Summary and conclusions

1.

Do individual differences in gradiency derive from differences in general cognitive function?

 Probably not.  Maybe speech perception operates on a different level than higher cognitive processes.

47

Summary and conclusions

1.

Do individual differences in gradiency derive from differences in general cognitive function?

 Probably not.  Maybe speech perception operates on a different level than higher cognitive processes. 2.

Are individual differences in gradiency linked to multiple cue use?

 Yes, more gradient listeners tend to rely more on the secondary cue (F0).  Better encoding of fine-grained detail (more gradiency) enables access to multiple cues.  And/or more gradient listeners commit later to a category.

48

Summary and conclusions

1.

Do individual differences in gradiency derive from differences in general cognitive function?

 Probably not.  Maybe speech perception operates on a different level than higher cognitive processes. 2.

Are individual differences in gradiency linked to multiple cue use?

 Yes, more gradient listeners tend to rely more on the secondary cue (F0).  Better encoding of fine-grained detail (more gradiency) enables access to multiple cues.  And/or more gradient listeners commit later to a category. 3.

In what way are these differences important for speech perception?

 More gradient listeners do a bit better (marginally) in perceiving speech in noise.  Gradiency is not all that bad - maybe good for some things.

49

Take home messages

1.

Gradiency indicates more accurate, true-to-the-signal perception.

50

Take home messages

1.

Gradiency indicates more accurate, true-to-the-signal perception.

2.

Some listeners are more gradient than others in categorizing phonemes.

51

Take home messages

1.

Gradiency indicates more accurate, true-to-the-signal perception.

2.

Some listeners are more gradient than others in categorizing phonemes.

3.

This gradiency may be a good thing.

52

Thank you!

53